Pharmacological Perioperative Brain Neuroprotection: a Qualitative Review of Randomized Clinical Trials

Home , Remacemide

British Journal of Anaesthesia 110 (S1): i113–i120 (2013) Advance Access publication 5 April 2013 . doi:10.1093/bja/aet059 Pharmacological perioperative brain neuroprotection: a qualitative review of randomized clinical trials

F. Bilotta1*, A. W. Gelb2, E. Stazi1, L. Titi1, F. P. Paoloni3 and G. Rosa1

1 Department of Anesthesiology, Critical Care and Pain Medicine, Section of Neuroanesthesia and Neurocritical Care, ‘Sapienza’ University of Rome, Rome, Italy 2 Department of Anesthesia and Perioperative Care, University of California, San Francisco, CA, USA 3 G.I.M.E.M.A. Onlus, Rome, Italy * Corresponding author. E-mail: [email protected]

Summary. Perioperative cerebral damage may be associated with surgery and anaesthesia. Editor’s key points Pharmacological perioperative neuroprotection is associated with conflicting results. In this † This review of 25 qualitative review of randomized controlled clinical trials on perioperative pharmacological Downloaded from randomized controlled brain neuroprotection, we report the effects of tested therapies on new postoperative clinical trials addresses neurological deficit, postoperative cognitive decline (POCD), and mortality rate. Studies perioperative were identified from Cochrane Central Register and MEDLINE and by hand-searching. Of pharmacological 5904 retrieved studies, 25 randomized trials met our inclusion criteria. Tested therapies + neuroprotection. were: lidocaine, thiopental, S( )-ketamine, propofol, nimodipine, GM1 ganglioside, lexipafant, glutamate/aspartate and xenon remacemide, atorvastatin, magnesium http://bja.oxfordjournals.org/ † Effects of various sulphate, erythropoietin, piracetam, rivastigmine, pegorgotein, and 17b-estradiol. The use pharmacological agents of atorvastatin and magnesium sulphate was associated with a lower incidence of new on neurological and postoperative neurological deficit. The use of lidocaine, ketamine, and magnesium sulphate cognitive dysfunctions, was associated with controversial results on POCD. The POCD did not differ between treated and mortality were patients and control group for other tested drugs (thiopental, propofol, nimodipine, reviewed. GM1 ganglioside, lexipafant, glutamate/aspartate, xenon, erythropoietin, remacemide, † Importantly, atorvastatin piracetam, rivastigmine, pegorgotein, and 17b-estradiol). None of the tested drugs was by Reni Rouncivell on June 15, 2013 and magnesium sulphate associated with a reduction in mortality rate. reduced the incidence of Drugs with various mechanisms of action have been tested over time; current evidence neurological deficit. suggests that pharmacological brain neuroprotection might reduce the incidence of new † All other agents remain postoperative neurological deficits and POCD, while no benefits on perioperative mortality controversial in their role. are described. Of importance from this review is the need for shared methodological approach when clinical studies on pharmacological neuroprotection are designed. Keywords: brain neuroprotection; ketamine; lidocaine; magnesium sulphate; perioperative cerebral damage; postoperative cognitive decline; perioperative stroke

Perioperative brain damage is among the most serious adverse hospitalization, impairs postoperative quality of life, and complications of surgery and anaesthesia, resulting in new increases perioperative costs.489One unresolved key question postoperative neurological deficits including transient ischae- is whether pharmacological strategies, independent of mech- mic attack (TIA), stroke, and postoperative cognitive decline anism of action, can effectively reduce the clinical impact of (POCD).12The incidence of new postoperative neurological perioperative brain damage. deficits ranges from 0.08% after general surgery to 5.2% The aim of this review of randomized controlled clinical after cardiac surgery.34The risk of perioperative stroke is trials (RCTs) on perioperative pharmacological brain neuro- increased in cardiovascular and neurovascular procedures protection is to assess the effects of the tested therapies and in patients with predisposing risk factors such as previous on the incidence of new postoperative neurological deficit, stroke, carotid stenosis, patent foramen ovale, atrial fibrilla- POCD, and mortality rate. tion, infective endocarditis, diabetes, renal failure, and older age (.62).4 –6 The incidence of POCD ranges from 28% to 100% after cardiac surgery and from 7% to 26% after non- Methods cardiac surgical procedures.78Perioperative brain damage To identify trials for inclusion in this review, a detailed remains a concern because it increases mortality, lengthens systematic research using Cochrane Central Register of

Controlled Trials and MEDLINE was performed. The terms New postoperative neurological deficit used are listed in Appendices 1 and 2. Only complete New postoperative neurological deficit was reported in 10 RCTs studies were included, not abstracts. that tested nine drugs: thiopental, GM1 ganglioside, nimodi- We identified RCTs that met the following criteria: (i) used pine, propofol, pegorgoteine, atorvastatin, magnesium sul- any pharmacological therapy for perioperative brain neuro- phate, 17b-estradiol, and ketamine.12 14 17 19 22 26 30 31 33 34 protection; (ii) evaluated neurological status before operation The incidence of new postoperative neurological deficit was and after operation; (iii) measured the cognitive status before lower in studies that tested atrovastatin and magnesium sul- operation and after operation using the same tests; and (iv) phate was associated with conflicting results for thiopental included adult patients (at least 18 yr of age with no upper and did not differ between treated patients and control limit) undergoing elective surgery. group for the other tested drugs (GM1 ganglioside, nimodipine, Two authors (F.B. and E.S.) independently screened and propofol, pegorgoteine, 17b-estradiol, and ketamine) (Table 1). assessed titles, abstracts, and the full-text papers. Details Atorvastatin, administered to 100 vascular surgery of study population, interventions, and outcomes were patients at a dose of 20 mg day21 for at least 15 days extracted using a standardized data extraction form that before operation and for a total of 45 days, was associated includes: eligibility and exclusion criteria, study design, dur- with a lower incidence of new postoperative neurological ation of follow-up, randomization, allocation concealment, deficits detected clinically daily, until discharge, and blinding, number and characteristics of patients, type of monthly, until the 6th month, and confirmed by Downloaded from surgery, drug dosage, and method of administration. neuroimaging.19 The outcome measures in this review are: Magnesium sulphate was tested in one RCT. One hundred and seventy-four cardiac surgery patients received a 780 mg Primary outcome measure: new postoperative neurological bolus dose during anaesthesia induction followed by 3169 deficit defined as stroke, with the appearance of symptoms

mg in a continuous infusion for 24 h. Assessment at 96 h http://bja.oxfordjournals.org/ and/or focal signs in the physical examination confirmed by after operation with WPNS showed a better composite computerized tomography (CT) imaging, or as a change in primary endpoint when compared with the 176 patients postoperative score from preoperative assessment with who served as the control group.31 neurological scales such as the National Institutes of Thiopental, administered by continuous infusion titrated Health Stroke Scale (NIHSS) and the Western perioperative to EEG isoelectricity, was tested in 482 patients undergoing neurologic scale (WPNS) at short-term (≤2 months) cardiac surgery in two RCTs.26 34 Both trials found no and/or long-term postoperative follow-up (.2 months). reduction in the incidence of new postoperative neurological Secondary outcome measures: POCD assessed as a 20% deficits. However, one of these trials, Nussmeier and by Reni Rouncivell on June 15, 2013 or greater deficit from baseline on postoperative testing colleagues,26 found at 10 days that when data on new or as decline .1 standard deviation in the postoperative neurological and psychiatric dysfunction were pooled for cognitive test scores, compared with the preoperative the 182 treated patients, the incidence was significantly score, graded with a battery of neuropsychological tests reduced (P,0.025). evaluating a broad array of cognitive domains as proposed by each study authors. Some investigators chose to also use other neurocognitive endpoints, in addition and for Postoperative cognitive deficits completeness, we have included these in the Results The POCD was evaluated in 24 RCTs that tested 16 drugs: lido- section as ‘neurocognitive performance’. caine in four studies, magnesium sulphate, thiopental, S(+)- Mortality. ketamine, propofol, nimodipine in two studies each, while GM1 ganglioside, lexipafant, glutamate/aspartate, xenon, erythropoietin, remacemide, piracetam, rivastigmine, pegor- Results gotein, and 17b-estradiol were tested in one study. 10 –18 20–34 Five thousand nine hundred and four studies were screened The incidence of POCD did not differ between treated patients and 5872 excluded as not randomized, duplicate, or irrele- and control group for tested drugs [thiopental, S(+)-ketamine, vant. Twenty-five RCTs—which included 3274 patients (age propofol, nimodipine, GM1 ganglioside, lexipafant, glutamate/ range 22–86 yr)—were retrieved and analysed (Fig. 1).10 – 34 aspartate, xenon, erythropoietin, remacemide, piracetam, riv- Characteristics of the studied population are summarized in astigmine, pegorgotein, and 17b-estradiol], and was asso- Table 1. Patients enrolled were of both sex with age ciated with conflicting results for lidocaine, ketamine, and ranging from 22 to 86 yr. Twenty-two of the 25 studies magnesium sulphate. were in cardiac surgery patients,10 –18 20 22 24 –34 two were Lidocaine was evaluated in four studies that enrolled a in vascular surgery,19 23 and one was in major abdominal total of 571 cardiac patients.16 24 28 32 Three evaluated and urological surgery.21 All studies, but one, provide short- POCD at short-term postoperative follow-up (9, 10 days, 6 term postoperative follow-up (≤2 months),10 – 34 10 studies weeks) and two at long-term postoperative follow-up (25 provide both short- and long-term (.2 months) follow-up,11 weeks and 6 months). The trials differ in time and dosing 12 16 19 22 25 27 30–32 and one trial provides only long- of the lidocaine infusion. In two studies, lidocaine was admi- term (.2 months) follow-up data.24 nistered as a 48 h continuous infusion, in one study as a 12 h i114 Perioperative neuroprotection BJA

Database search 5904

Other sources 0 MEDLINE 1983 National research register and metaregister of controlled trials EMBASE 1914 PubMED 1009

CENTRAL 998

National research register and metaregister of controlled trials

Excluded 5872 Irrelevant/not RCT/duplicate

Full paper review 32 Downloaded from Excluded 7 Included trials 25

Observational study 2 Trials testing lidocaine 4

Post hoc analysis 2 Trial testing thiopental 2 Non-pharmachological study 2 Trial testing S(+)-ketamine 2 http://bja.oxfordjournals.org/

Pre- and postoperative evaluation 1 Trail testing propofol 2

not consistent Trial testing nimodipine 2

Trial testing magnesium sulfate 2

Trial testing GM1 ganglioside 1

Trials testing lexipafant 1

Trial testing glut/asp 1 by Reni Rouncivell on June 15, 2013

Trial testing xenon 1

Trial testing remacemide 1

Trial testing atorvastatin 1

Trial testing erythropoietin 1

Trial testing piracetam 1

Trial testing rivastigmine 1

Trial testing pergogotein 1

Trial testing 17b-estradiol 1

Fig 1 Study selection for inclusion in meta-analysis. continuous infusion, while in one study, lidocaine infusion ketamine was injected as an i.v. bolus (0.5 mg kg21) during was stopped at the end of surgery. The first two studies anesthesia induction, while in the other, it was given as showed a significant benefit in treated patients, while the bolus (2.5 mg kg21) during anesthesia induction followed latter failed to confirm a neuroprotective benefit.16 24 28 32 by continuous infusion (125 mgkg21 min21) until the end Furthermore, the authors of the latter comment that the of the surgical procedure. Benefits in terms of reduction of failure may be related to a too short an infusion period, an POCD at 1 week after surgery were limited to the trial that excessively high dose, or having included diabetic patients tested the lower dose of ketamine bolus injection (0.5 mg who might be at increased risk of neurological injury or a kg21) during anesthesia induction.14 different sensitivity to lidocaine.32 Magnesium sulphate was tested for the prevention of Ketamine was evaluated in two trials that enrolled a total POCD in two RCTs with conflicting results.23 31 In a trial of of 172 patients undergoing cardiac surgery.14 25 The trials 350 cardiac patients randomized to control or to receive a differ in timing and dosing of ketamine. In one trial, 780 mg bolus dose during anaesthesia induction followed

i115 i116 BJA

Table 1 Characteristics of randomized controlled trials included in the qualitative review

Study ID Surgery Drug studied Dose of drug studied No. of pts No. of pts Postoperative New postoperative POCD receiving receiving follow-up neurological deﬁcit drug placebo Mitchell and colleagues16 Cardiac Lidocaine 1 mg kg21 in bolus; 4 mg min21 for 1 h; 2 32 32 10 days; 10 NA P¼0.025 at short-term mg min21 for 2 h; 1 mg min21 for 46 h weeks; 6 months follow-up; P¼NS at long-term follow-up Wang and colleagues28 Cardiac Lidocaine 1.5 mg kg21 in bolus; 4 mg min21 until the 57 61 9 days NA P¼0.028 end of operation Mitchell and colleagues24 Cardiac Lidocaine 1 mg kg21 bolus; 2 mg min21 for 2 h; 1 mg 81 77 10; 25 weeks NA P¼NS at early and min21 for 10 h long-term follow-up Mathew and colleagues32 Cardiac Lidocaine 1 mg kg21 in bolus; 4 mg min21 for 1 h; 2 114 127 6 weeks; 1 yr NA P¼NS at early and mg min21 for 46 h long-term follow-up Nussmeier and colleagues26 Cardiac Thiopental 50–100 mg until EEG became isolelectric 89 93 5 days P¼NS P¼NS; P.0.025 for neurological and neuropsychological complications Zaidan and colleagues34 Cardiac Thiopental To create burst suppression 149 151 2; 5 days P¼NS P¼NS Nagels and colleagues25 Cardiac Ketamine 2.5 mg kg21 bolus; 125 mgkg21 min21 58 62 1; 10 week NA P¼NS Hudetz and colleagues14 Cardiac Ketamine 0.5 mg kg21 in bolus 26 26 1 week NA P,0.01 Roach and colleagues33 Cardiac Propofol Infusion just burst suppression 109 116 1–2; 5–7; 50–70 P¼NS P¼NS at early and days long-term follow-up Kanbak and colleagues15 Cardiac Propofol 6; 3 mg kg21 h21 10 10 3; 6 days NA P¼NS Forsman and colleagues11 Cardiac Nimodipine 0.5 mgkg21 min21 20 19 5 days; 6 months NA P¼NS Legault and colleagues22 Cardiac Nimodipine 60 mg before surgery; 30 mg every 6 h for 5 75 74 1 week; 1; 6 P¼NS P¼NS days after surgery months Bhudia and colleagues31 Cardiac Magnesium 780 mg over 15 min; 3160 mg over 24 h 174 176 1; 4 days; 3 P¼NS; better Magnesium sulphate did sulphate months composite primary not inﬂuence endpoint at 96 h neuropsychological evaluated with WPNS function 3 months after operation Mack and colleagues23 Vascular Magnesium (I) 10 g 500 ml21; (II) 18 g 500 ml21; (III) 20 55 53 1 day NA Low-dose group (I–II): sulphate g 500 ml21 P¼0.01; high-dose group (III): P¼NS Grieco and colleagues12 Cardiac GM1 300 mg×2 19 11 1 week; 6 months P¼NS P¼NS ganglioside Taggart and colleagues27 Cardiac Lexipafant 0.4; 4 mg h21 100 50 5 days; 3 months NA P¼NS Erol and colleagues10 Cardiac Glutamate/ 15 mmol litre21 35 35 3 days NA P¼NS aspartate- enriched

crystalloid Bilotta cardioplegia Ho¨cker and colleagues21 Major non Xenon 60% in O 56 58 1; 6; 30 days NA P¼NS

2 al et cardiac

Downloaded from from Downloaded http://bja.oxfordjournals.org/ by Reni Rouncivell on June 15, 2013 15, June on Rouncivell Reni by eiprtv neuroprotection Perioperative

Arrowsmith and Cardiac Remacemide 25–50–100–150 mg 18 day; 150 ×4/day 87 84 8 weeks NA P¼NS; the remacemide colleagues18 group showed signiﬁcantly superior performance in three tests over the control group Durazzo and colleagues19 Vascular Atorvastatine 20 mg day21 for 45 days 50 50 Until the P¼0.18; treated NA discharge; 6 patients had a months signiﬁcant decrease in the rate of cardiac events Haljan and colleagues13 Cardiac Erythropoietin 375; 750; 1500 units kg21 24 8 On day of NA P¼NS at early and discharge; 2 long-term follow-up months Holinski and colleagues20 Cardiac Piracetam 12 g in bolus 60 60 3 days NA P.0.0005 cognitive decline was less in all subtests of the treated patients except attention Gamberini and colleagues29 Cardiac Rivastigmine Three 1.5 mg doses per day every 8 h 59 61 Daily for 6 days NA P¼NS starting on the evening preceding the operation and continuing until the evening of the 6th postoperative day Butterworth and Cardiac Pegorgotein 5000 UI kg21 i.v. bolus; 2000 UI kg21 i.v. 45 22 5–7 days; 4–6 P¼NS. No patients P¼NS at early and colleagues17 bolus weeks demonstrated a long-term follow-up periopertaive stroke Hogue and colleagues30 Cardiac 17b-estradiol 7.6 mg in transdermal patch; 0.08 ng kg21 86 88 4–6 weeks; 6 P¼NS P¼NS at early and min21 infusion during surgery months long-term follow-up BJA

i117

Downloaded from from Downloaded http://bja.oxfordjournals.org/ by Reni Rouncivell on June 15, 2013 15, June on Rouncivell Reni by BJA Bilotta et al. by 3169 mg in a continuous infusion for 24 h, no reduction in damage than other high-risk patients. Perioperative neuro- POCD was found.31 Magnesium sulphate given to 49 vascular logical deficits have been detected in up to 80% of in-hospital surgery patients treated with a bolus during anaesthesia patients after cardiac surgery and it is still present in as many induction, dose range 2–4 g, and followed by a continuous as 42% 3–5 yr later. In non-cardiac surgery, perioperative 24 h infusion, dose range 8–16 g, when compared with the neurological deficit/cognitive decline is much less frequent 43 control patients, was associated with a lower incidence and can be detected in 26% of patients in short-term follow-up of POCD at the first postoperative day in the 13 patients and in 10% in long-term follow-up.145The higher incidence who received less than a total of 10 g of magnesium probably reflects specific factors unique to cardiac surgery, sulphate (bolus+infusion).23 so that neuroprotective strategies pertinent to those patients Finally, the use of remacemide and piracetam, although may not be appropriate for other high-risk patients.42 48 not effective in reducing POCD, yielded a better postoperative Studies in cardiac surgery should be viewed as hypothesis gen- ‘neurocognitive performance’.18 20 erating but needing to be confirmed in the different surgical settings. Mortality No consensus exists on the best neuropsychometric tests Mortality was evaluated in 16 RCTs that tested 12 drugs: for detecting and quantifying neurological damage and lidocaine, thiopental, propofol, ketamine, remacemide, nimo- POCD. Neither is there agreement on the optimal timing for postoperative testing, both for research and for day-to-day dipine, atorvastatin, magnesium sulphate, xenon, rivastig- 49 – 51 Downloaded from mine, pegorgotein, and 17b-estradiol.11 16 – 19 21 22 24 25 28 – 34 clinical use. Cognition is the result of activities in mul- No statistically significant differences in the mortality rate tiple complex, distributed, and interacting neuronal circuits among treated and untreated patients were found. that underlie specific information processing functions. Of interest, one RCT that tested nimodipine was prema- Therefore, comprehensive neuropsychological assessment requires a battery of tests assessing a variety of cognitive turely suspended because of a higher mortality rate in 1 http://bja.oxfordjournals.org/ patients randomized to receive nimodipine compared with domains. In this review, we included only studies that those in the placebo group.22 used consistent pre- and postoperative evaluation methods for the neurological and cognitive status in order to minimize the risk of bias related to matching patients. For the same Discussion reason, we also excluded patients with acute brain damage The main finding of this review is that in some experimental (i.e. trauma or acute subarachnoid haemorrhage).17 20 The paradigms, a few drugs appear to potentially have a neuro- study that tested rivastigmine for perioperative neuroprotec-

protective effect. However, the methodological inconsisten- tion was included in this meta-analysis, despite having post- by Reni Rouncivell on June 15, 2013 cies and weakness, and the small number of studies do not operative delirium as primary endpoint because it presented allow any firm conclusions. cognitive dysfunction evaluated before and after operation The studies we analysed have many methodological according to our inclusion criteria as secondary endpoint.29 discrepancies: rationale for drug selection, dosing, timing, A potential limitation of this review is related to the sep- length of administration, type of surgery, and type and timing arate analysis for new postoperative neurological deficits of neurocognitive testing. The drugs tested for pharmacological and POCD. We decided to independently analyse these two brain neuroprotection have many differing mechanisms of complications because current research practice categorizes action and their dosing and timing were selected according to new focal neurological deficits (type 1 outcome) separately preclinical studies associated with positive results.35–41 from POCD, confusion, agitation, memory loss, and seizures However, in animal studies, neurological outcome is usually when they are without evidence of focal injury (type 2 based on fairlysimple tasks, whereas in humans, neurocognitive outcome).252A second limitation is the heterogeneity of sur- testing evaluates complex neuronal functions and tasks.35–37 gical procedures studied. Of interest when the same drug, 42–45 These difference might, in part, explain the lack of positive magnesium sulphate, was tested in cardiac and non-cardiac effects when pharmacological brain neuroprotection clinical patients, both studies reported a neuroprotective effect.23 31 trials are designedbasedon animalstudies. A retrospective ana- This single example though does not validate the extrapola- lysis on the effects of statin and b-blockers in patients who had tion of results from cardiac surgery to other types of surgery. undergone coronary artery bypass graft, not included in this In conclusion, based on single studies, pharmacological review because of the study design, demonstrates protective neuroprotection might provide benefits in terms of a reduc- effects on postoperative stroke and provides interesting insights tion in new postoperative neurological deficits or POCD, but on methodological criteria for drug selection.46 no effects on mortality were reported. Almost all studies to Although our review was not intended to focus primarily on date have been done in cardiac surgery patients, where the neuroprotection in cardiac surgery, the majority of selected more general applicability is unclear. Future studies need to studies (22 out of the 25) were accomplished in this clinical include a broader range of relevant clinical scenarios using setting. It is controversial if the perioperative brain damage a wider consensus on the methodological approaches, that takes place during cardiac surgery represents an appropri- including timing and dosing of drug administration, patient ate study model for other types of surgery.24347These patients selection, and perioperative neurological and cognitive have a markedly higher incidence of perioperative brain testing. i118 Perioperative neuroprotection BJA

Acknowledgements 18 Arrowsmith JE, Harrison DM, Newman SP. Neuroprotection of the brain during cardiopulmonary bypass: a randomized trial of The authors are grateful to Dr William J. Mack for having remacemide during coronary artery bypass in 171 patients. provided the details of the study necessary for this Stroke 1998; 29: 2357–62 meta-analysis. The authors are grateful to Prof. Albert Sau- 19 Durazzo AE, Machado FS, Ikeoka DT, et al. Reduction in cardiovas- berman and Prof. Deepak Tempe for their thoughtful insights. cular events after vascular surgery with atorvastatin: a randomized trial. J Vasc Surg 2004; 39: 967–76 References 20 Holinski S, Claus B, Alaaraj N, et al. Cerebroprotective effect of piracetam in patients undergoing coronary bypass surgery. Med Sci 1 Silverstein JH, Timberger M, Reich DL, Uysal S. Central nervous Monit 2008; 14: PI53–7 system dysfunction after noncardiac surgery and anesthesia in 21 Ho¨cker J, Stapelfeldt C, Leiendecker J, et al. Postoperative neuro- the elderly. Anesthesiology 2007; 106: 622–8 cognitive dysfunction in elderly patients after xenon versus 2 Carrascal Y, Guerrero AL. Neurological damage related to cardiac propofol anesthesia for major noncardiac surgery: a double- surgery. Neurologist 2010; 16: 152–64 blinded randomized controlled pilot study. Anesthesiology 2009; 3 Ng JL, Chant MT, Gelb AW. Perioperative stroke in noncardiac, 110: 1068–76 nonneurosurgical surgery. Anesthesiology 2011; 115: 879–90 22 Legault C, Furberg CD, Lynne E, et al. Nomidipine neuroprotection 4 Selim M. Perioperative stroke. N Engl J Med 2007; 356: 706–13 in cardiac valve replacement. Stroke 1996; 27: 593–8 5 Newman MF, Kirchner JL, Phillips-Bute B, Neurological Outcome 23 Mack WJ, Kellner CP, Sahlein DH, et al. Intraoperative magnesium Research Group and the Cardiothoracic Anesthesiology Research infusion during carotid endarterectomy: a double-blind placebo- Downloaded from Endeavors Investigators. Longitudinal assessment of neurocogni- controlled trial. J Neurosurg 2009; 110: 961–7 tive function after coronary-artery bypass surgery. N Engl J Med 24 Mitchell SJ, Merry AF, Frampton C, et al. Cerebral protection by 2001; 344: 395–402 lidocaine during cardiac operations: a follow-up study. Ann 6 Mashour GA, Shannks AM, Kheterpal S. Perioperative stroke and Thorac Surg 2009; 87: 820–5 associated mortality after noncardiac, nonneurologic surgery. 25 Nagels W, Demeyere R, Van Hemelrijck J, Vandenbussche E, Anesthesiology 2011; 114: 1289–96 Gijbels K, Vandermeersch E. Evaluation of the neuroprotection http://bja.oxfordjournals.org/ 7 McKhann GM, Goldsborough MA, Borowicz LM Jr,et al. Cognitive effects of S(+)-ketamine during open-heart surgery. Anesth outcome after coronary artery bypass: a one-year prospective Analg 2004; 98: 1595–603 study. Ann Thorac Surg 1997; 63: 510–5 26 Nussmeier NA, Arlund C, Slogoff S. Neuropsychiatric complica- 8 Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term post- tions after cardiopulmonary bypass: cerebral protection by a bar- operative cognitive dysfunction in the elderly: ISPOCD1 study. biturate. Anesthesiology 1986; 64: 140–2 Lancet 1998; 351: 861 27 Taggart DP, Browne SM, Wade DT, Halligan PW. Neuroprotection 9 Bateman BT, Schumacher HC, Wang S, Shaeﬁ S, Berman MF. Peri- during cardiac surgery: a randomised trial of a platelet activating operative acute ischemic stroke in noncardiac and nonvascular

factor antagonist. Heart 2003; 89: 897–900 by Reni Rouncivell on June 15, 2013 surgery: incidence, risk factors and outcomes. Anesthesiology 28 Wang D, Wu X, Li J, Xiao F, Liu X, Meng M. The effect of lidocaine 2009; 110: 231–8 early postoperative cognitive dysfunction after coronary artery 10 Erol DD, Ibis HA. Glutamate/aspartate supplementation during bypass surgery. Anesth Analg 2002; 95: 1134–41 cardiopulmonary bypass: effect on postoperative neurocognitive 29 Gamberini M, Bolliger D, Lurati Buse GA, et al. Rivastigmine for the function. Heart Surg Forum 2008; 11: E17–20 prevention of postoperative delirium in elderly patients undergo- 11 Forsman M, Tubylewicz Olsnes B, Semb G, Steen PA. Effects of ing elective cardiac surgery—a randomized controlled trial. Crit nimodipine on cerebral blood flow and neuropsychological Care Med 2009; 37: 1762–8 outcome after cardiac surgery. Br J Anaesth 1990; 65: 514–20 30 Hogue CW Jr, Freedland K, Hershey T, et al. Neurocognitive out- 12 Grieco G, D’Hollosy M, Culliford AT, Saran J. Evaluating neuropro- comes are not improved by 17b-estradiol in postmenopausal tective agents for clinical anti-ischemic benefit using neurological women undergoing cardiac surgery. Stroke 2007; 38: 2048–54 and neuropsychological changes after cardiac surgery under 31 Bhudia SK, Cosgrove DM, Naugle RI, et al. Magnesium as a neuro- cardiopulmonary bypass. Stroke 1996; 27: 858–74 protectant in cardiac surgery: a randomized clinical trial. J Thorac 13 Haljan G, Maitland A, Buchan A, et al. The erythropoietin neuro- Cardiovasc Surg 2006; 131: 853–61 protective effect assessment in CABG surgery (TENPEAKS): a ran- 32 Mathew JP, Mackensen GB, Phillips-Bute B, et al. Randomized, domized, double-blind, placebo controlled, proof-of-concept double-blinded, placebo controlled study of neuroprotection clinical trial. Stroke 2009; 40: 2769–75 with lidocaine in cardiac surgery. Stroke 2009; 40: 880–7 14 Hudetz JA, Iqbal Z, Gandhi SD, et al. Ketamine attenuates 33 Roach GW, Newman MF, Murkin JM, et al. Multicenter Study of post-operative cognitive dysfunction after cardiac surgery. Acta Perioperative Ischemia (MsSPI). Ineffectiveness of burst suppres- Anaesthesiol Scand 2009; 53: 864–72 sion therapy in mitigating perioperative cerebrovascular dysfunc- 15 Kanbak M, Saricaoglu F, Avci A, Ocal T, Koray Z, Aypar U. Propofol tion. Anesthesiology 1999; 99: 1255–64 offers no advantage over isoflurane anesthesia for cerebral pro- 34 Zaidan JR, Klochany A, Martin WM, Ziegler JS, Harless DM, tection during cardiopulmonary bypass: a preliminary study of Andrews RB. Effect of thiopental on neurologic outcome follow- b S-100 protein levels. Can J Anaesth 2004; 51: 712–7 ing coronary artery bypass grafting. Anesthesiology 1991; 74: 16 Mitchell SJ, Pellett O, Gorman DF. Cerebral protection by lidocaine 406–11 during cardiac operations. Ann Thor Surg 1999; 67: 1117–24 35 Bilotta F, Doronzio A, Stazi E, et al. Postoperative cognitive 17 Butterworth J, Legault C, Stump DA, et al. A randomized, blinded dysfunction: toward the Alzheimer’s disease pathomechanism trial of the antioxidant pegorgotein: no reduction in neuro- hypothesis. J Alzheimers Dis 2010; 22:81–9 psychological deficits, inotropic drug support, or myocardial is- 36 Mantz J, Degos V, Laigle C. Recent advance in pharmacologic chemia after coronary artery bypass surgery. J Cardiothorac neuroprotection. Eur J Anaesthesiol 2010; 27:6–10 Vasc Anesth 1999; 13: 690–4

i119 BJA Bilotta et al.

37 Hall R, Murdoch J. Brain protection: physiological and pharmaco- #5 MeSH descriptor Cardiac Surgical Procedures explode logical considerations. Part II: The pharmacology of brain protec- all trees tion. Can J Anaesth 1990; 37: 762–77 #6 MeSH descriptor Neurosurgery explode all trees 38 Gelb AW, Bayona NA, Wilson JX, Cechetto DF. Propofol anesthesia #7 MeSH descriptor Foramen Ovale, Patent explode all compared to awake reduces infarct size in rats. Anesthesiology trees 2002; 96: 1183–90 #8 MeSH descriptor Extracorporeal Circulation explode all 39 Yang Q, Dong H, Deng J, et al. Sevoflurane preconditioning induces neuroprotection through reactive oxygen species- trees mediated up-regulation of antioxidant enzymes in rats. Anesth #9 MeSH descriptor Atrial Fibrillation explode all trees Analg 2011; 112: 931–7 #10 MeSH descriptor Endocarditis explode all trees 40 Koerner IP, Brambrink AM. Brain protection by anesthetic agents. #11 MeSH descriptor Carotid Stenosis explode all trees Curr Opin Anaesthesiol 2006; 19: 481–6 #12 MeSH descriptor Anesthesia, General, this term only 41 Schifilliti D, Grasso G, Conti A, Fodale V. Anaesthetic-related neu- #13 MeSH descriptor Anesthesia, Local explode all trees roprotection: intravenous or inhalational agents? CNS Drugs 2010; #14 MeSH descriptor Anesthetics, Local, this term only 24: 893–907 #15 Barbiturat* or Propofol or Isoflurane or Lidocaine or 42 Brown WR, Moody DM, Challa VR. Cerebral fat embolism from car- Lignocaine or Nimodipine or GM1 ganglioside or Remace- diopulmonary bypass. J Neuropathol Exp Neurol 1999; 58: 109–19 mide or (adrenergic receptor antagonist*) or ketamine or 43 Grogan K, Stearns J, Hogue CW. Brain protection in cardiac Thiopental or Thiopentone or Carotid surgery. Anesthesiol Clin 2008; 26: 521–38 Downloaded from endoarterectom* or [(Cardi* or Open heart) near Surg*] or 44 Rees K, Beranek-Stanley M, Burke M, Ebrahim S. Hypothermia to reduce neurological damage following coronary artery bypass Coronary artery bypass or Neurosurg* or Patent Foramen surgery. Cochrane Database Syst Rev 2001: CD002138. Ovale or Extracorporeal circulation or Atrial fibrillation or doi:10.1002/14651858.CD002138 Infective endocarditis or Carotid stenosis 45 Todd MM, Hindman BJ, Clarke WR, Toner JC. Intraoperative #16 (#4 OR #5 OR #6 OR #7 OR # OR #9 OR #10 OR #11 OR http://bja.oxfordjournals.org/ Hypothermia for Aneurysm Surgery Trial (IHAST) Investigators. #12 OR #13 OR #14 OR #15) Mild intraoperative hypothermia during surgery for intracranial #17 (#3 AND #16) aneurysm. N Engl J Med 2005; 352: 135–45 46 Bouchard D, Carrier M, Demers P, et al. Statin in combination with b-blocker therapy reduces postoperative stroke after coronary artery bypass graft surgery. Ann Thorac Surg 2011; 91: 654–9 Appendix 2: MEDLINE 47 Mackensen GB, Gelb AW. Postoperative cognitive deficits: more (1) {[(cerebral or brain) protect*] or neuro?protect*}.mp. questions than answers. Eur J Anaesthesiol 2004; 21:85–8 (2) Neuroprotective Agents/

48 Motallebzadeh R, Jahangiri M. The effect of the dynamic air (3) 1 or 2 by Reni Rouncivell on June 15, 2013 bubble trap on cerebral microemboli and S100 beta. J Thor (4) {(An?esthetic* adj3 Drug*) or Barbiturat* or Propofol or Cardiovasc Surg 2004; 128: 154 Isoﬂurane or Lidocaine or Lignocaine or Nimodipine or 49 Monk TG, Weldon C, Garvan CW, et al. Predictors of cognitive dys- GM1 ganglioside or Remacemide or (adrenergic recep- function after major noncardiac surgery. Anesthesiology 2008; tor antagonist* or ketamine) or Thiopental or Thiopen- 108:18–30 tone or Carotid endoarterectom* or [(Cardi* or Open 50 Gladstone D, Black S, Hakim A; Heart and Stroke Foundation of Ontario Centre of Excellence in Stroke Recovery. Towards heart) adj3 Surg*] or Coronary artery bypass or Neuro- wisdom from failure: lessons from neuroprotective stroke trials surg* or Patent Foramen Ovale or Extracorporeal circu- and new therapeutic directions. Stroke 2002; 33: 2123–36 lation or Atrial ﬁbrillation or Infective endocarditis or 51 Murkin J, Newman S, Stump D, Blumenthal JA. Statement of con- Carotid stenosis}.mp. sensus on assessment of neurobehavioral outcomes after cardiac (5) exp Endarterectomy, Carotid/ or exp Cardiac Surgical surgery. Ann Thorac Surg 1995; 59: 1289–95 Procedures/ or exp Neurosurgery/ or exp Foramen 52 Roach GW, Kanchuger M, Mangano CM, et al. Adverse cerebral Ovale, Patent/ or exp Extracorporeal Circulation/ or outcomes after coronary bypass surgery. N Engl J Med 1996; exp Atrial Fibrillation/ or exp Endocarditis/ or exp 335: 1857–63 Carotid Stenosis/ or exp Anesthesia, General/ or exp Anesthesia, Local/ or exp Anesthetics, Local/ Appendix 1: CENTRAL, The Cochrane Library (6) 4 or 5 #1 MeSH descriptor Neuroprotective Agents explode all (7) 3 and 6 trees (8) [(randomized controlled trial or controlled clinical #2 ((cerebral or brain) near protect*) or neuroprotect* trial).pt. or randomized.ab. or placebo.ab. or drug ther- #3 (#1 OR #2) apy.fs. or randomly.ab. or trial.ab. or groups.ab.] not #4 MeSH descriptor Endarterectomy, Carotid explode all [animals not (humans and animals)].sh. trees (9) 7 and 8

Handling editor: R. P. Mahajan

i120